The STRING database in 2017: quality-controlled protein–protein association networks, made broadly accessible

The STRING database collects and integrates known and predicted protein–protein associations across a large number of organisms, aiming for a system-wide view of cellular function. Its associations include both direct physical interactions and indirect functional ones, as long as they are specific and biologically meaningful. In addition to reassessing experimental interaction data and importing curated pathways and complexes, STRING derives predictions from systematic co-expression analysis, detection of shared selective signals across genomes, automated text-mining of the scientific literature, and computational transfer of interaction knowledge between organisms based on gene orthology.

In version 10.5 the largest changes concern data dissemination: the web frontend was completely redesigned to reduce dependency on outdated browser technologies, and the database can now be queried from inside the widely used Cytoscape software framework. Further improvements include automated background analysis of user inputs for functional enrichments and streamlined download options, making the resource more accessible and useful for interpreting protein interaction networks.